megyn: do you think dick morris is right? he says latinos increased their share by 1.5%. he says if you use that same model for this election and assume you will have the same turnouts from blacks, latinos and young kids, you will be off in your results. >> we just don't know the answer. that is, if all groups are depressed in their turnout as seems likely. the black or latino share will be about the same. young people probably will go down. they are showing less interests. but we can't say with specificity what dick is asserting. megyn: what are you supposed to do? >> both of us -- political pollsters screen and likeliness to vote, interest, intensity and where you are. we have in 2004 a split country. we had about an 8-point advantage for democrats in 2000. i can't imagine it being as good or better. if it is, republicans are going down across the board. >> we try to use a variety of different scenarios. even party, slight democratic advantage. bigger democratic advantage then look at differen different altes and advice accordingly. megyn: is there a finger on the poll?